PROJECT SUMMARY/ABSTRACT Translational cancer research requires robust preclinical models to most effectively investigate the underlying biology of disease and develop new therapeutics. While all models are imperfect, it is essential to understand the degree by which each model system (e.g., cell culture, xenograft) recapitulates specific molecular and functional characteristics of human tumors. This may be particularly relevant for studying altered metabolism, a hallmark of cancer, as even subtle changes to the environment can greatly impact the metabolic phenotype of a tumor. Moreover, as cancer metabolism is tightly regulated by oncogenic signaling, the diversity of molecular alterations within a given malignancy may elicit unique metabolic characteristics; which, may greatly influence metabolic pathway dependencies for tumor proliferation and growth. To best determine the fidelity of preclinical models in preserving the metabolic features of human cancer, this requires cross-comparing matched patient tissue and preclinical models across tumors with various genetic alterations. However, such a comprehensive investigation has yet to be undertaken. This proposal will perform an integrated metabolic and molecular characterization of matched human tumors, direct-from-patient orthotopic xenografts (GliomaPDOX), and cell lines from patients with glioblastoma (GBM) ? one of the most lethal human malignancies that also reside within the unique brain metabolic milieu. In Aim 1, stable isotope-labeled metabolic tracing and liquid chromatography- mass spectrometry (LC-MS) will be used to cross-compare the metabolic phenotypes of prospectively matched GBM patient tumors, GliomaPDOX, and cell lines to determine the metabolic characteristics that are preserved and/or lost from patient to preclinical model. Aim 2 proposes to determine, in genetically diverse preclinical GliomaPDOX models, whether specific metabolic phenotypes align with distinct molecular signatures. Finally, in Aim 3, in vivo genetic knockdown experiments will be performed to assess whether measured metabolic phenotypes represent targetable dependencies for GliomaPDOX growth, invasion, and survival. Collectively, the studies proposed in this application will provide critical insight into the translatability of preclinical GBM models for studying tumor metabolism; which, may ultimately have important implications for developing new therapeutics against metabolic dependencies in GBM, and potentially, other malignancies.